The purpose of the proposed research is to synthesize a series of peptide derivatives of the antitumor drug actinomycin D (AMD). They will be tested in prodrug and targeting strategies aimed at improving tumor specificity. The title compounds will consist of short peptides (3-5 amino acids) attached amide-wise to the AMD chromophoric amino group. A method has been developed to synthesize such novel compounds, which, unlike AMD itself, do not bind to DNA. A peptidyl-actinomycin will act as a prodrug if cleaved to AMD by an enzyme found at higher levels in tumor than normal tissue. Proposed examples are Glu (Delta)-Gly-Pro- (AMD) and D-Val-Leu-lys-Val-Pro-(AMD) which should be cleaved by delta-glutamyl transferase (GGT) and plasmin respectively. Such cleavage would afford the dipeptidyl-actinomycin which will spontaneously cyclize to the diketopiperazine and AMD. This would circumvent the reported difficulty with some peptidyl- drugs that they are not substrates for the enzyme. GGT is increased in human tumors of the liver, colon and breast; plasmin is associated with many solid tumors. The GGT-activated prodrugs will be tested in various cell lines with high or low GGT levels. They and the plasmin-activated prodrugs will be submitted to NCI for test in tumor-bearing animals. For targeting, conjugation of other peptidyl-AMD's (designed for cleavage to AMD by lysosomal enzymes) with monoclonal antibodies to specific tumor cells will be undertaken. To increase the AMD/antibody ratio, conjugates will also be prepared by linking antibody to human serum albumin after conjugating the latter with many molecules of peptidyl-AMD. These targeting approaches will be tested initially in vitro in appropriate human tumor cell lines and subsequently in tumor bearing animals. In vitro testing will include cell viability assays in human leukemia cell lines after exposure to the peptidyl-AMD-antibody conjugates.